Table 3.

Percent of total days supplied attributable to Drugs to Avoid in the Elderly (DAE) before and after Part D

	Unadjusted		Adjusted Impact of Part D		Comparison of Adjusted Impact of Part D
	Pre-Part D (%)	Post-Part D (%)	Adjusted Odds Ratio (95% Confidence Interval)	p-value	Ratio of Odds Ratios (95% Confidence Interval)	p-value
No coverage	3.01	1.98	0.68 (0.59-0.78)	<0.0001	0.84 (0.72-0.98)	0.03
$150 cap	2.08	1.78	0.84 (0.75-0.95)	0.004	1.04 (0.91-1.20)	0.553
$350 cap	1.80	1.45	0.82 (0.78-0.86)	<0.0001	1.01 (0.93-1.10)	0.794
No Cap	1.72	1.34	0.81 (0.75-0.87)	<0.0001	reference group

Adjusted for age, sex, prospective risk score, census block group-level data on race, education and income using GEE negative binomial model.

Pre-Part D time period is January 1, 2004-December 31, 2005. Post-Part D time period is January 1, 2006-December 31, 2007. DAE measure excludes benzodiazepines which were not covered by Part D